Tankyrase 1 and tankyrase 2 are essential but redundant for mouse embryonic development

Abstract

Tankyrases are proteins with poly(ADP-ribose) polymerase activity. Human tankyrases post-translationally modify multiple proteins involved in processes including maintenance of telomere length, sister telomere association, and trafficking of glut4-containing vesicles. To date, however, little is known about in vivo functions for tankyrases. We recently reported that body size was significantly reduced in mice deficient for tankyrase 2, but that these mice otherwise appeared developmentally normal. In the present study, we report generation of tankyrase 1-deficient and tankyrase 1 and 2 double-deficient mice, and use of these mutant strains to systematically assess candidate functions of tankyrase 1 and tankyrase 2 in vivo. No defects were observed in development, telomere length maintenance, or cell cycle regulation in tankyrase 1 or tankyrase 2 knockout mice. In contrast to viability and normal development of mice singly deficient in either tankyrase, deficiency in both tankyrase 1 and tankyrase 2 results in embryonic lethality by day 10, indicating that there is substantial redundancy between tankyrase 1 and tankyrase 2, but that tankyrase function is essential for embryonic development.

Figure 1. Generation of TANK1-deficient mice by gene targeting.

(A) Gene targeting strategy and restriction map of TANK1 gene. Filled boxes indicate exons; labeled boxes indicate neomycin (neo) resistance or herpesvirus thymidine kinase (tk) genes; and arrows indicate loxP sites. (B, C) Southern blot analysis of ES cell DNA. The 4.2 kb and 1.5 kb bands represents the germ line alleles and 6.0 kb and 2.5 kb bands represent the targeted alleles when BamHI/XbaI enzymes were used (B). The 12 kb and 1.3 kb bands represent the germ line alleles, and 13.8 kb and 0.8 kb bands represent the targeted alleles when EcoRV was used(c). (D) PCR analysis for the conditional TANK1 knockout mouse genotype. The 340- and 240-bp PCR products represent the wild-type and floxed alleles, respectively. (E) PCR analysis for the constitutive TANK1 knockout mouse genotype. The 340- and 200-bp PCR products represented the wild-type and knockout alleles, respectively.

Figure 2. Expression of TANK1 and TANK1a in WT and TANK1^−/− mouse tissues.

All RNA samples were serially diluted as indicated for PCR amplification and analysis. Actin cDNA was used as an RT-PCR loading control in all experiments. (A) RT-PCR analysis for TANK1 mRNA expression in various tissues of WT and TANK1^−/− (KO) mice as indicated. 5′-TANK1 RT-PCR products represent upstream cDNA of TANK1. (B) RT-PCR analysis for TANK1 mRNA expression in WT and TANK1^−/− (KO) testis as indicated. 5′-TANK1 and 3′-TANK1 RT-PCR products represent upstream and downstream cDNAs of TANK1; and TANK1a RT-PCR products are specific for TANK1a. (C) RT-PCR analysis for TANK1a mRNA expression in various tissues of WT mice as indicated. TANK1a RT-PCR products represent cDNA of TANK1a. (D, E) Western blot analysis was used to determine tankyrase 1 and 1a expression in thymus, testis and spleen of WT and TANK1^−/− (KO) mice with 762 (anti-SAM, D) and 376 (anti-HPS, E) antibodies. TANK1 indicates tankyrase 1 protein, TANK1x indicates a possible degraded tankyrase 1 protein, and TANK1a indicates tankyrase 1a protein.

Figure 3. Genomic structure, mRNA and protein product for TANK1 and TANK1a.

Figure 4. Telomere length was not altered in TANK1 knockout mice.

Spleen cells were isolated from C57BL/6, TANK1^+/+ (n = 7) and TANK1^−/− (n = 7) mice, and relative telomere length was determined by Flow-FISH. The FITC fluorescent signal of the cell-binding telomeric probe was converted to arbitrary units of molecule equivalents of soluble fluorescence (MSEF). The average of fluorescent intensities from each mouse was normalized to that of a C57BL/6 mouse (defined as 100). The relative telomere length of each strain of mice is plotted.

Figure 5. Cell cycle progression is equivalent in wild-type, TANK1^−/−, and TANK2^−/− mice.

(A) Spleen cells from TANK1^−/−, TANK2^−/− mice and their littermates were stimulated with IL-2, ConA and LPS for 48 hours, then fixed and stained with PI. The results shown are representative of 3 independent analyses. The percentages of G1, S and G2/M phase are indicated. (B) Spleen cells from TANK1^−/−, TANK2^−/− mice and their littermates were stimulated with IL-2, ConA and LPS for 48 hours, gamma irradiated at the 24 hour time point, then fixed and stained with PI. The results shown are representative of 3 independent analyses. The percentages of G1, S and G2/M phase are indicated.

Figure 6. The body weights of tankyrase 1 knockout mice are normal while the body weights of tankyrase 2 knockout mice are reduced.

(A) Body weights in male TANK1^+/+ (n = 8), TANK1^−/− (n = 8), TANK2+/+ (n = 5) and TANK2−/− (n = 5) mice. (B) Body weights in female TANK1^+/+ (n = 5), TANK1^−/− (n = 5), TANK2^+/+ (n = 4) and TANK2^−/− (n = 4) mice.

Figure 7. The sensitivity of insulin-stimulated glucose uptake in tankyrase 1 knockout WAT is normal while the sensitivity of insulin-stimulated glucose uptake in tankyrase 2 knockout WAT is decreased.

(A) Assay for insulin-responsive glucose uptake in TANK1^+/+and TANK1^−/− WAT cells. The results shown represent three independent experiment, each using two male TANK1^+/+ and 2 male TANK1^−/− mice. (B) Assay for insulin-responsive glucose uptake in TANK2^+/+and TANK2^−/− WAT cells. The results shown represent five independent experiments, each using two male TANK2^+/+ and 2 male TANK2^−/− mice.

Figure 8. Abnormal phenotypes in TANK1^−/−.TANK2^−/− mouse embryos and placentas.

(A, C) wild-type and (B, D) TANK1^−/−.TANK2^−/− E10.5 embryos and placentas. Arrow points to umbilical cord. Histological sections of wild-type (E) and TANK1^−/−.TANK2^−/− (F) E10.5 embryos. Transverse sections of E10.5 wild-type (G) and TANK1^−/−.TANK2^−/− (H) placentas. The box area in (G) and (H) are shown at a higher magnification in (I) and (J), respectively.